Creation and validation of a tool to measure the real population risk of speeding

Speeding is a significant contributor to the road toll. The risks associated with speeding at high and low levels are well understood, as are a range of speed management programs such as speed limits and speed cameras. Speed surveys have been used to help quantify the population distribution of speeding and to evaluate the impact of road safety interventions, especially since crashes are, statistically, rare events. Crash-based evaluations necessarily take time and are subject to high levels of random variation. In contrast, speed surveys can potentially be used to demonstrate the immediate impact of road safety interventions. The results of speed surveys are available long before the associated crash data has been collected and analysed. However, the achieved changes in the distribution of vehicle speeds need to be translated into changes in the likely number and severity of crashes. The utility of speed survey data is diminished, however, through the use of crude summary statistics such as mean, 85 percentile and percentage of drivers exceeding the speed limit. Recently D’Elia et al. (2008), Doecke et al. (2011), Gavin et al. (2010) have combined known casualty crash risk estimates with speed surveys to further our understanding of how the population level of risk of speeding can be described and evaluated through the use of speed surveys. This paper documents the development of a tool which translates speeding identified in speed surveys into a risk measure. The tool is designed to be used both with annual general network speed surveys and with speed surveys used to evaluate the impact of road safety interventions such as enforcement programs. The tool has been validated using speed survey and crash data from evaluations of the trial introduction of the 50 km/h urban speed limit, a speed limit reduction on the Great Western Highway and an evaluation of NSW fixed speed cameras.